Encoding and concealing information using deep learning

Abstract

Aspects of the subject disclosure may include, for example, a method for training a deep learning model that includes encoding a content item; generating a blended image by combining a background image and the encoded content; decoding the blended image to generate decoded content corresponding to the content item; and defining or specifying a loss function related to the deep learning model. The method also includes determining values of training parameters for the deep learning model to minimize the loss function, thereby obtaining a trained deep learning model. The method also includes an information concealing procedure using the trained deep learning model to conceal user content by encoding the user content and blending the encoded user content with a user-selected image; the information concealing procedure is substantially independent of the user-selected image. Other embodiments are disclosed.

What is claimed is: 
     
       1. A device comprising:
 a processing system including a processor; and 
 a memory that stores executable instructions that, when executed by the processing system, facilitate performance of operations, the operations comprising: 
 calculating a first calculated loss determined by a first matrix representing a first background image and a second matrix representing a first blended image, the first blended image comprising the first background image and first encoded content based on a first content item; 
 decoding the first blended image to generate decoded content; 
 calculating a second calculated loss determined by a third matrix representing the first content item and a fourth matrix representing the decoded content; 
 calculating a loss function in accordance with the first calculated loss and the second calculated loss to generate a model; and 
 performing a procedure using the model, wherein the model is trained using parameters to minimize the loss function, resulting in a trained model, 
 wherein the procedure comprises: 
 encoding a second content item to generate second encoded content; and 
 generating a second blended image by combining a second background image and the second encoded content, wherein the procedure is substantially independent of the second background image. 
 
     
     
       2. The device of  claim 1 , wherein the first encoded content comprises an encoded text image. 
     
     
       3. The device of  claim 1 , wherein the first content item and the second content item comprise a same content. 
     
     
       4. The device of  claim 1 , wherein the decoding the first blended image comprises extracting the first encoded content from the first blended image and decoding the first encoded content, wherein the decoding is performed using a portion of the trained model. 
     
     
       5. The device of  claim 1 , wherein the operations further comprise generating the first blended image by performing an element-wise combination of the first matrix and an encoded content matrix representing the first encoded content. 
     
     
       6. The device of  claim 5 , wherein the first calculated loss corresponds to a distortion of the first background image due to generating the first blended image, and wherein the second calculated loss corresponds to a reconstruction error determined by comparing the first content item with the decoded content. 
     
     
       7. The device of  claim 1 , wherein the first background image is randomly selected. 
     
     
       8. A method comprising:
 calculating, by a processing system including a processor a first calculated loss determined by a first matrix representing a first background image and a second matrix representing a first blended image, the first blended image comprising the first background image and first encoded content based on a first content item; 
 decoding, by the processing system, the first blended image to generate decoded content; 
 calculating, by the processing system, a second calculated loss determined by a third matrix representing the first content item and a fourth matrix representing the decoded content; 
 generating, by the processing system, a model based on the first calculated loss and the second calculated loss; and 
 performing, by the processing system, a procedure using the model, wherein the model is trained using parameters to minimize a loss function based on the first calculated loss and the second calculated loss, resulting in a trained model, wherein the procedure comprises:
 encoding a second content item to generate second encoded content; and 
 generating a second blended image by combining a second background image and the second encoded content, wherein the procedure is substantially independent of the second background image. 
 
 
     
     
       9. The method of  claim 8 , further comprising:
 encoding, by the processing system, a text message using a trained autoencoder model: and 
 generating, by the processing system, the first encoded content including the encoded text image. 
 
     
     
       10. The method of  claim 8 , wherein the trained model comprises a deep learning model. 
     
     
       11. The method of  claim 10 , wherein the decoding the first blended image comprises extracting the first encoded content from the first blended image and decoding the first encoded content, wherein the decoding is performed using a portion of the deep learning model. 
     
     
       12. The method of  claim 8 , further comprising generating, by the processing system, the first blended image by performing an element-wise combination of the first matrix and an encoded content matrix representing the first encoded content. 
     
     
       13. The method of  claim 8 , wherein the first background image is randomly selected. 
     
     
       14. A non-transitory machine-readable medium comprising executable instructions that, when executed by a processing system including a processor, facilitate performance of operations, the operations comprising:
 calculating a first calculated loss determined by a first matrix representing a first background image and a second matrix representing a first blended image, the first blended image comprising the first background image and first encoded content based on a first content item; 
 decoding the first blended image to generate decoded content; 
 calculating a second calculated loss determined by a third matrix representing the first content item and a fourth matrix representing the decoded content; 
 calculating a loss function in accordance with the first calculated loss and the second calculated loss to generate a model, wherein the model is trained using parameters to reduce the loss function, resulting in a trained model; and 
 performing a procedure using the trained model, the procedure comprising:
 encoding a second content item to generate second encoded content; and 
 generating a second blended image by combining a second background image and the second encoded content. 
 
 
     
     
       15. The non-transitory machine-readable medium of  claim 14 , wherein the first blended image further comprises a hidden text. 
     
     
       16. The non-transitory machine-readable medium of  claim 15 , wherein the procedure is substantially independent of the second background image. 
     
     
       17. The non-transitory machine-readable medium of  claim 14 , wherein the decoding the first blended image comprises extracting the first encoded content from the first blended image and decoding the first encoded content, wherein the decoding is performed using a portion of the trained model. 
     
     
       18. The non-transitory machine-readable medium of  claim 14 , wherein the operations further comprise generating the first blended image by performing an element-wise combination of the first matrix and an encoded content matrix representing the first encoded content. 
     
     
       19. The non-transitory machine-readable medium of  claim 18 , wherein the first calculated loss corresponds to a distortion of the first background image due to generating the first blended image, and wherein the second calculated loss corresponds to a reconstruction error determined by comparing the first content item with the decoded content. 
     
     
       20. The non-transitory machine-readable medium of  claim 14 , wherein the first background image is randomly selected.